Samsung has announced a new feature for its Bixby mobile assistant that creates an AI-generated copy of a user’s voice to answer calls. The new “Custom Voice Creator” lets users record different sentences for Bixby to analyze and create an AI-generated copy of their tone and voice. Unfortunately, the feature is currently only available in Korean on Galaxy S23 series devices.

The feature is designed to let you use your AI-generated voice to answer phone calls for you when you’re unable to. For instance, if you can’t answer a phone call, you could use the new feature to type out a sentence that will be read out to the caller in your voice. Samsung says it plans to make the feature compatible with other Samsung capabilities beyond phone calls in the future.

Samsung also announced that its Bixby Text Call feature, which transforms voice calls into written text, now supports calls in English. The feature initially launched last year in Korean and lets users answer calls by typing a message that Bixby will then read out loud to the caller. Text Call can also transcribe what the caller says. The feature can be accessed on Galaxy S23 devices, the Z Fold 4 and the Z Flip 4.

In addition, the company says it’s rolling the ability for users to customize Bixby’s wake-up phrase into whatever they want. The voice assistant can now also play music depending on the type of exercise that is playing on Samsung Health.

“Now, Bixby can better understand intent and process follow-up requests by understanding context and associating words previously used in interactions,” the company wrote in a blog post. “For example, Bixby users can first launch a workout on Samsung Health and then ask Bixby to play music that best suits that exercise by saying “Play music for this workout.”

Samsung also announced that it has expanded the things that Bixby can do offline, such as setting a timer, taking a screenshot or turning on the flashlight.

Samsung debuts a Bixby feature that creates an AI-generated copy of a user’s voice to answer calls by Aisha Malik originally published on TechCrunch